The development of valvular heart disease in patients with carcinoid syndrome is thought to be related to the secretion of vasoactive substances by a tumor. We sought to identify modifiable risk factors for the development of carcinoid heart disease because this may help define strategies to attenuate the disease process. Two hundred fifty-two patients with carcinoid syndrome were prospectively followed with serial echocardiograms at 6-month intervals. Clinical characteristics, biochemical markers, and radiologic markers were measured at set intervals. An echocardiographic scoring system was applied. Patients were defined as having progression of carcinoid heart disease if the echocardiographic score increased by ≥25%. After a median follow-up of 29 months, 44 patients developed carcinoid heart disease or had progression of existing valvular dysfunction. At time of progression of carcinoid heart disease compared to the previous 6 months, there was a significant increase in median levels of 5-hydroxyindoleacetic acid (5-HIAA; 791 vs 460.5 μmol/24 hours) and flushing episodes (4.5 vs 2 episodes per day). Independent predictors of the development or progression of carcinoid heart disease were a 5-HIAA level ≥300 μmol/24 hours and ≥3 episodes of flushing per day. 5-HIAA levels of ≥300 to 599, 600 to 899, and >900 μmol/24 hours conferred 2.74, 3.16, and 3.40 times the risk of progression of carcinoid heart disease, respectively. In conclusion, a 5-HIAA level ≥300 μmol/24 hours and ≥3 flushing episodes per day are predictors of the development or progression of carcinoid heart disease.
The development of carcinoid heart disease is classically characterized by the development of predominately right-sided valvular dysfunction. Progression of valvular dysfunction eventually leads to right heart failure and significantly impairs survival. Mechanisms for development of valvular dysfunction are incompletely understood. Increased levels of 5-hydroxytryptamine (5-HT) are thought to be a key mediator in the development of carcinoid heart disease. However >50% of patients with increased 5-HT do not develop carcinoid heart disease. Therefore, other cofactors may also be involved. The purpose of the present study was to prospectively identify clinical, metabolic, and tumor characteristics that predict patients with carcinoid syndrome who are at risk of developing or progression of carcinoid heart disease.
Methods
We conducted an observational cohort study. Patients with histologically proved metastatic carcinoid tumor of midgut origin and carcinoid syndrome were consecutively and prospectively recruited from the neuroendocrine tumor clinic at the Royal Free Hospital (London, United Kingdom) from April 2006 through January 2010. All patients gave written informed consent for the study. The study was approved by the institution’s local ethics committee.
All patients were evaluated at 4- to 6-month intervals in the neuroendocrine tumor clinic. Symptoms of carcinoid syndrome were assessed with a symptom score. Patients were classified as having symptomatic progression if there was a >50% increase in the median daily number of flushing episodes or bowel motions compared to the previous clinic visit or baseline (i.e., first study) visit.
Urinary 5-hydroxyindoleacetic acid (5-HIAA) and plasma chromogranin A (CgA) were measured at baseline and at each clinical evaluation (4- to 6-month intervals). CgA was measured using a DAKO chromogranin A enzyme-linked immunosorbent assay (DAKO A/S, Glostrup, Denmark). Patients were classified as having biochemical progression if there was a 50% increase of biochemical markers compared to the previous clinic visit or baseline value.
Baseline triple-phase computed tomography of the thorax, abdomen, and pelvis was undertaken. Restaging computed tomographic (CT) scan was performed at 4- to 6-month intervals. All CT scans were reported by radiologists with expertise in neuroendocrine tumors. Radiologic assessment was based on Response Evaluation Criteria in Solid Tumor (RECIST) criteria. CT scans were compared to the previous scan and the baseline CT scan.
Individual patient therapies were based on symptom status, biochemical and radiologic data, tumor histology, and pattern/extent of disease in accordance with international guidelines. Somatostatin analogues were used as first-line therapy for control of symptoms related to hormonal hypersecretion. A strategy of dose escalation was initially tried followed by a decrease in the interval between doses. For patients with uncontrolled symptoms despite somatostatin analogue, other therapies were used including interferon, radionuclide therapy, or hepatic embolization. Routine antidiarrheal medications were not prescribed.
Baseline comprehensive 2-dimensional transthoracic echocardiography was performed using an Acuson Sequoia C512 (Siemens Medical Systems, Mountain View, California). Repeat echocardiograms were performed at 6-month intervals or sooner if clinically indicated. Valve morphology and function were evaluated in several views using a standardized imaging protocol including the use of pulse, continuous-wave, and color Doppler. Valvular regurgitation was quantified into grades based on an integrated approach including valve structure/function, semiquantitative parameters (color flow jet area/width), quantitative parameters (vena contracta, proximal isovelocity surface area), and supportive signs in accordance with American Society of Echocardiography (ASE) guidelines. Valve stenosis was quantified into grades based on parameters recommended by ASE guidelines. Tricuspid stenosis was graded (mean gradient across valve) as mild (2 to 5 mm Hg), moderate (5 to 8 mm Hg), or severe (>8 mm Hg). Right and left ventricular functions and sizes were assessed and calculated according to ASE guidelines. Right ventricular diameter was measured at the midcavity of the right ventricle in the apical 4-chamber view. Right ventricular function was measured using fractional area change. Left ventricular ejection fraction was calculated by the modified Simpson method from apical 4- and 2-chamber views.
Carcinoid heart disease was defined as the presence of characteristic thickening, decreased excursion, and/or retraction of valvular leaflets (with associated evidence of valvular stenosis or regurgitation) in the absence of other causes.
Quantification of severity was performed by scoring each valve individually ( Table 1 ). Right ventricular size and function were graded according to ASE guidelines. Scores of 0, 1, 2, and 3 were assigned to normal, mild, moderate, and severe right ventricular dilatation and dysfunction, respectively. The sum of scores for each valve plus the right ventricle score produced a total score.
| Score | ||||
|---|---|---|---|---|
| 0 | 1 | 2 | 3 | |
| Leaflet thickness (mm) | <3 | ≥3–<4 | ≥4–<5 | ≥5 |
| Leaflet excursion | normal | ≤75% but >50% of normal | ≤50% but >25% of normal | ≤25% of normal or fixed |
| Leaflet retraction | normal | mild | moderate | severe |
| Valve stenosis | normal | mild | moderate | severe |
| Valve regurgitation | normal | mild | moderate | severe |
All echocardiograms were reviewed and scored by a reader (S.B.) with advanced training in echocardiography and blinded to clinical, biochemical, and radiological data. A second reader (J.D.) reviewed and scored a sample of 50 patients (20%) to calculate interobserver variability. A sample of 25 echocardiograms was reread to calculate intraobserver variability. Patients were defined as having progression of carcinoid heart disease if their score increased by ≥25%. Patients were defined as having developed carcinoid heart disease if features of carcinoid heart disease were not present on previous echocardiogram.
Descriptive statistics were used for patients and tumor characteristics at study entry. Data are expressed as median and interquartile range or number and percentage. Mann–Whitney U test was used to compare differences between continuous variables and chi-square test was used to compare categorical variables. For comparison of paired variables the Wilcoxon test was used.
A time-to-event analysis was undertaken using a Poisson regression model. Univariate Poisson regression analysis was performed to identify risk factors for progression of carcinoid heart disease. Multivariate Poisson analysis was performed to identify independent predictors of progression of carcinoid heart disease. To account for changes in variables during the study period, the following variables were considered as time-updated variables: 5-HIAA, CgA, and episodes of diarrhea and flushing. Therefore, the effect of the most recent level of the variable on the outcome was considered. All tests of significance were 2-sided. A probability value <0.05 was considered statistically significant. Statistical analysis was performed using StatsDirect 2.5.7 (StatsDirect, Altrincham, Cheshire, United Kingdom).
Results
Two hundred fifty-two patients were recruited. Median follow-up was 29 months (interquartile range 24 to 36). At baseline 41 patients had carcinoid heart disease. Forty-four patients developed carcinoid heart disease during the study period or had progression of existing carcinoid heart disease (15 patients developed and 29 patients had progression). Thirty-four patients died during the study period (13 patients had carcinoid heart disease). Twenty-two patients underwent cardiac valve replacement surgery. Five patients were lost to follow-up during the study period.
Baseline demographics are presented in Table 2 . There were no significant differences between the 2 groups. There were significantly higher baseline levels of 5-HIAA and CgA in those patients with progression of carcinoid heart disease.
| Progression of Carcinoid Heart Disease | No Progression of Carcinoid Heart Disease | p Value | |
|---|---|---|---|
| (n = 44) | (n = 208) | ||
| Age (years) | 60 (55–67) | 63 (55–69) | 0.35 |
| Women | 22 (50%) | 110 (53%) | 0.86 |
| Number of echocardiograms | 5 (4–6) | 5 (3–6) | 0.09 |
| Follow-up (months) | 32 (24–36) | 29 (24–34) | 0.10 |
| Duration of diagnosis of carcinoid tumor (months) | 5 (3–6) | 4 (3–7) | 0.67 |
| Tumor grade | |||
| Intermediate | 4 (9%) | 11 (5%) | 0.31 |
| Low | 40 (91%) | 197 (95%) | 0.31 |
| Presence of liver metastases | 42 (96%) | 186 (89%) | 0.27 |
| Therapy | |||
| Somatostatin analogue | 42 (96%) | 178 (86%) | 0.08 |
| Dose of somatostatin analogue (mg) | |||
| Octreotide LAR | 30 (20–30) | 30 (20–30) | 0.89 |
| Lanreotide autogel | 90 (90–120) | 90 (90–120) | 0.45 |
| Duration of somatostatin therapy (months) | 30 (18–44) | 27 (18–41) | 0.85 |
| Interferon | 4 (9%) | 10 (5%) | 0.28 |
| Chemotherapy (FCiSt) | 4 (9%) | 12 (6%) | 0.49 |
| Targeted radionuclide | 17 (39%) | 56 (27%) | 0.14 |
| Surgical resection | 15 (34%) | 100 (48%) | 0.10 |
| Liver metastases resection | 1 (2%) | 2 (1%) | 0.50 |
| 5-hydroxyindolacetic acid (μmol/24 hours) | 465 (244–1,324) | 46.5 (0–205) | <0.001 |
| Chromogranin A (pmol/L) | 990 (323–1,000) | 206 (72–529) | <0.001 |
In patients with development or progression of carcinoid heart disease there was a significant increase in levels of 5-HIAA and median episodes of flushing at time of progression compared to the previous 6 months. There were no significant changes in median levels of CgA or episodes of diarrhea during the same period ( Table 3 ). Nineteen of the 44 patients (43%) with progression of carcinoid heart disease had an increase in the size of liver metastases (progressive disease by RECIST criteria on computed tomogram).
| Progressive Carcinoid Heart Disease | No Progression of Carcinoid Heart Disease | |||||
|---|---|---|---|---|---|---|
| (n = 44) | (n = 208) | |||||
| At Time of Progression | Previous 6 Months | Significance | Baseline | Peak | Significance | |
| 5-Hydroxyindolacetic acid (μmol/24 hours) | 791 (581–1,084.5) | 460.5 (309–948.5) | 0.001 | 52.5 (0–205) | 55 (0–205) | 0.09 |
| Chromogranin A (pmol/L) | 1,000 (428.5–1,000) | 976.5 (323–1,000) | 0.17 | 206 (72–520) | 209.5 (68.5–561.5) | 0.08 |
| Episodes of flushing/24 hours | 4.5 (4–6) | 2.0 (1–3) | <0.001 | 1 (0–1) | 1 (0–1) | 0.45 |
| Episodes of diarrhea/24 hours | 2 (2–3) | 3 (2–4.5) | 0.09 | 2 (1–3) | 2 (0.5–3) | 0.45 |
In patients without development or progression of carcinoid heart disease there was no significant change in median levels (baseline compared to peak) of 5-HIAA, CgA, or episodes of diarrhea or flushing during the study period ( Table 3 ). Twenty-five of 208 patients (12%) without progression of carcinoid heart disease had an increase in the size of liver metastases (progressive disease by RECIST criteria on computed tomogram).
Incidence rate ratios for development or progression of carcinoid heart disease are listed in Table 4 . Flushing of >3 episodes over 24 hours, progressive symptoms (flushing), 5-HIAA levels ≥300 to 599, ≥600 to 899, and ≥900 μmol/24 hours, progression of 5-HIAA, presence of >5 liver metastases and progression of liver metastases were associated with progression of carcinoid heart disease.
| Parameter | Incidence Rate Ratio | 95% Confidence Interval | Significance |
|---|---|---|---|
| Age | 0.99 | 0.97–1.02 | 0.49 |
| Women | 0.87 | 0.48–1.57 | 0.64 |
| Diarrhea >5 stools/24 hours | 1.95 | 0.94–4.05 | 0.08 |
| Progressive symptoms (diarrhea) | 1.27 | 0.53–3.01 | 0.58 |
| Flushing >3 episodes/24 hours | 26.87 | 10.59–68.18 | <0.001 |
| Progressive symptoms (Flushing) | 21.46 | 9.98–46.20 | <0.001 |
| 5-Hydroxyindolacetic acid (μmol/24 hours) | |||
| <300 | 0.01 | 0.00–0.09 | <0.001 |
| 300–599 | 3.26 | 1.76–6.02 | <0.001 |
| 600–899 | 4.51 | 2.36–8.61 | <0.001 |
| >900 | 5.17 | 2.71–9.89 | <0.001 |
| Progressive biochemical markers (5-hydroxyindolacetic acid) | 10.07 | 5.56–18.23 | <0.001 |
| Chromogranin A >500 pmol/L | 1.73 | 0.95–3.15 | 0.07 |
| Progressive biochemical markers (chromogranin A) | 1.82 | 0.84–3.91 | 0.13 |
| >5 Liver metastases | 4.19 | 1.95–9.01 | <0.001 |
| Progressive disease (radiologic) | 2.85 | 1.57–5.17 | <0.001 |
| Tumor grade—intermediate | 0.55 | 0.20–1.55 | 0.26 |
| Therapy | |||
| Chemotherapy | 1.39 | 0.50–3.89 | 0.53 |
| Somatostatin analogue | 1.75 | 0.54–5.63 | 0.35 |
| Targeted therapy | 1.63 | 0.90–2.96 | 0.11 |
| Interferon | 0.98 | 0.30–3.16 | 0.97 |
| Surgical resection | 0.54 | 0.29–1.03 | 0.06 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
